arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2307.04420cs.DCcs.AI
收藏

FedDCT:无线网络中一种动态跨层联邦学习框架

FedDCT: A Dynamic Cross-Tier Federated Learning Framework in Wireless Networks

Guangxi Normal University(广西师范大学);School of Computer Science and Engineering, Guangxi Normal University(广西师范大学计算机科学与工程学院)

Youquan Xian, Xiaoyun Gan, Chuanjian Yao, Dongcheng Li, Peng Wang, Peng Liu, Ying Zhao

AI总结:

本文提出FedDCT框架,通过动态分层和跨层设备选择算法缓解无线网络中资源异构与滞后设备问题,显著减少收敛时间并提升收敛准确率。

摘要:

联邦学习(FL)作为一种保护隐私的机器学习范式,在不暴露本地数据的情况下跨设备训练全局模型。然而,无线网络中的资源异构性和不可避免的滞后设备严重影响了联邦学习训练的效率和准确性。本文提出了一种新颖的动态跨层联邦学习框架(FedDCT)。首先,我们设计了一种动态分层策略,根据设备的响应时间将其动态划分到不同层级,并为每个层级分配特定的超时阈值,以减少单轮训练时间。然后,我们提出了一种跨层设备选择算法,选择响应迅速且有利于模型收敛的设备,以提高收敛效率和准确性。实验结果表明,在无线网络下,所提出的方法优于基线方法,收敛时间平均减少了54.7%,收敛准确率平均提高了1.83%。

英文摘要:

Federated Learning (FL), as a privacy-preserving machine learning paradigm, trains a global model across devices without exposing local data. However, resource heterogeneity and inevitable stragglers in wireless networks severely impact the efficiency and accuracy of FL training. In this paper, we propose a novel Dynamic Cross-Tier Federated Learning framework (FedDCT). Firstly, we design a dynamic tiering strategy that dynamically partitions devices into different tiers based on their response times and assigns specific timeout thresholds to each tier to reduce single-round training time. Then, we propose a cross-tier device selection algorithm that selects devices that respond quickly and are conducive to model convergence to improve convergence efficiency and accuracy. Experimental results demonstrate that the proposed approach under wireless networks outperforms the baseline approach, with an average reduction of 54.7\% in convergence time and an average improvement of 1.83\% in convergence accuracy.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑